Abstract Insulin resistance is a major cause of chronic diseases including type 2 diabetes (T2D), heart attacks, strokes and cancer. Only one class of medications, thiazolidinediones (TZDs), specifically targets insulin resistance primarily by activating the transcription factor PPARG in adipocytes. While TZDs have proven clinically effective in preventing T2D, heart attacks and strokes, serious side effects have limited their clinical use. New therapeutic targets to combat insulin resistance are needed. In theory, the gene expression changes caused by TZD-treatment could be mined for novel insulin-sensitizing effectors, but TZDs alter the expression of hundreds of genes. These must be sifted by functional investigation for causal effectors versus merely correlated or toxic bystanders. Standard laboratory-based functional investigation requires invasive physiological measurements in model systems that are difficult to scale. Even when a potential insulin sensitizing effector gene is validated in the lab, credentialing its relevance to human insulin sensitivity necessitates drug development and human trials, another poorly scalable process that usually results in failure for lack of efficacy. However, the recent accumulation of genome sequences in large, clinically characterized populations has revealed that nature has performed countless human trials in the form of millions of naturally occurring, protein-altering genetic variants scattered throughput almost every gene in the genome. The key to unlocking both these opportunities: 1) identifying novel candidate genes from TZD treatment and 2) leveraging nature?s clinical trials for assessing therapeutic potential, are high-throughput functional assays. In this application, we propose to utilize a newly developed massively parallel adipocyte differentiation/ lipid accumulation assay in an integrative genomic approach to: Aim 1: Rapidly sort TZD-altered genes for likelihood of being insulin sensitizing effectors and Aim 2: Credential already identified and novel candidates for therapeutic potential in humans using data from 100,000 sequenced individuals. This work will produce a systematic dissection of the therapeutic effect of TZDs to identify novel insulin sensitivity genes and directly assess the therapeutic potential of modulating function in humans for four of these genes.